Richard Rovner, MathWorks
As markets accelerate, globalization extends, and smart systems expand into countless aspects of our lives, Model-Based Design has become an essential approach to systems engineering and development. This talk highlights recent applications of Model-Based Design, exposes the critical underlying technologies based on MATLAB and Simulink, overviews new product enhancements from MathWorks, and touches on the extension of Model-Based Design into academia, ensuring the growth of the next generation of engineers and scientists.
Sundar Umamaheswaran, MathWorks
This presentation shows how you can use MathWorks solutions for technical computing and algorithm development to process and analyse vibration data. Using measured vibration data represented as a time series, we build an algorithm to extract meaningful characteristics of interest such as undamped natural frequencies and damping ratios. The presentation discusses the complete technical computing workflow through the following stages:
- Importing and exploring data interactively
- Developing data analysis algorithms
- Using optimisation and curve-fitting techniques to extract meaningful characteristics of interest from data
- Automatically generating reports for documentation
- Automating the complete workflow
This session will be helpful for engineers and scientists working in the areas of data analysis, NVH, and postprocessing of acquired field data.
Chris Aden, MathWorks
Software-defined radios include reconfigurable subsystems for changing carrier frequency, receiver bandwidth, sampling rates, and digital baseband signal processing. This reconfigurability is realized in the radio frequency (RF) front end and associated analog electronics, as well as in the FPGA/DSP elements, which implement intermediate frequency (IF) and baseband digital signal processing algorithms. In this presentation, system simulation is used to prototype a wireless transceiver for a variety of communication links and operating conditions. The system-level model includes all of the RF and analog mixed-signal (AMS) subsystems that are present on a reconfigurable RFIC chip as well as the digital baseband code that is used in both the transmitter and the receiver. The simulation results are compared with a hardware-in-the-loop test using an evaluation board that includes the RFIC chip and a Xilinx® Virtex®-5 FPGA.
Permanent magnet synchronous motors (PMSMs) play a vital role in automotive and industrial automation applications. However, the development of PMSM controllers remains a complex and often expensive engineering task. To reduce development time and cost, engineers have turned to Model-Based Design, in which the controller is modelled as part of an executable specification, designed with simulation, and implemented using automatic code generation. In this session, MathWorks engineers discuss the benefits of Model-Based Design and demonstrate the design flow through a case study targeting TI's C2000 processors.
Introducing a Development Workflow from Concept to Realisation for First-Time Users of MathWorks Tools
Pradeep Nanjappa, MathWorks
Engineers and scientists across various industries and in numerous applications use MathWorks tools for technical computing and Model-Based Design to enable system modelling, simulation, verification, and implementation. If you are a new user and are unsure of where to get started, or if you have questions about the capabilities and benefits of the various tools, this session is for you. Using a video processing application example, we show how you can apply MATLAB, Simulink, and related tools from the initial concept stage all the way to implementation.
Design, Modelling, and Simulation Track
Puneet Kumar, MathWorks
System objects enable users to efficiently model and simulate dynamic systems by using the MATLAB language. Additionally, stream data processing capabilities enable the efficient processing of long data sets with limited memory usage. System objects, which are object-oriented implementations of time-based and data-driven algorithms, data-access routines, and visualisation tools, offer a consistent coding style that is easy to learn and is completely integrated into the MATLAB language. Using MATLAB Coder™, you can automatically generate standalone C code incorporating fixed-point or floating-point data types from systems modelled using System objects. Attendees in this session will discover how the System objects in DSP System Toolbox™, Communications System Toolbox™, and Computer Vision System Toolbox™ can be used to enable the design, analysis, and simulation of complex digital signal processing, communications, and video processing systems within MATLAB.
The session Implementation of Signal and Video Processing Systems Using MATLAB Coder details the implementation of System objects in C code.
Steve Miller, MathWorks
Batteries are the heart of hybrid electric vehicles (HEVs), unmanned vehicles, uninterruptable power supplies (UPSs), and many other modern-day electrical systems. An accurate battery model is essential to system-level design and analysis for many battery-based systems. However, battery models are complex and difficult to parameterize to match real-world data and require a refined level of fidelity to achieve meaningful simulation results. In this presentation, MathWorks experts develop a battery model and use test data to automatically tune the model parameters using MathWorks products. We show how you can use Simscape™ to implement the nonlinear equations of the equivalent circuit components of the battery model and Simulink Design Optimization™ to increase model accuracy by using test data to calibrate physical parameters.
Jasvinder Singh Mangat, LRDE, Bangalore
Synthetic aperture radar (SAR) systems achieve fine resolution in the cross-range coordinate by synthesizing a long array and spending more time on target; the resolution in the range coordinate is limited by the bandwidth of transmitted pulse. As the resolution requirement becomes finer, it becomes mandatory to transmit and process higher-bandwidth pulses. To meet these requirements, wide-bandwidth linear frequency modulated (LFM) waveforms are generated. To digitally generate these waveforms, samples are generated at much higher rate, reaching to the theoretical limits of the currently available programmable devices, such as FPGAs.
In this presentation, we discuss the various methodologies for generating wide-bandwidth waveforms and how MATLAB and Simulink were used in the initial simulation and development of an optimum wide-bandwidth waveform generation technique. The hardware has been realized with this technique, and MATLAB was used to fine-tune the algorithm and test the module against the specifications.
Chris Aden, MathWorks
In the realm of high-speed digital design, the critical issue of signal integrity poses increasing challenges to design engineers. In signal integrity applications where the ultimate goal is to ensure high-speed data transmission, designers often need to build transmission line models manually from measured data to test I/O circuit designs. This session shows how you can use MathWorks tools to model transmission lines as rational functions. This type of behavioural model is faster and more accurate and provides greater insight into transmission line characteristics than traditional alternatives such as inverse fast Fourier transforms (IFFTs). The session also shows how MathWorks tools can help in building impairment models of backplanes and PCB traces using modified rational functions, enforcing passivity and causality on S-parameters, performing backplane analysis, and designing pre-emphasis and equaliser coefficients for a certain data rate.
Tom Erkkinen, MathWorks
An increasing number of engineers and scientists are using Model-Based Design for developing algorithms and systems. In Model-Based Design, automatic code generation from models is a key part of the implementation stage, but it is also used throughout design and development. A control design engineer uses code generation for rapid prototyping and hardware-in-the-loop (HIL) testing, whereas an embedded software engineer working in a production program uses it to deploy code onto an embedded microcontroller of an ECU or on a PLC controller used in an industrial application.
In this presentation, we show how you can use code generation products such as Simulink Coder™, Embedded Coder™, and Simulink PLC Coder™ for rapid prototyping, production code generation, and HIL testing. We demonstrate code generation steps including model checking, floating-point to fixed-point conversion, algorithm export and scheduling, target optimisation, and inclusion of legacy code. Code generation support for industry standards such as MISRA-C, AUTOSAR, IEC 61508, and ISO 26262 is also described.
Tabrez Khan, MathWorks
In traditional system development processes, high-level representations of algorithms are rewritten in C/C++ and other languages for implementation on microcontrollers and DSPs or for integration into larger software packages. Implementations in C/C++ are also used for verification or simulation acceleration purposes. Translating MATLAB algorithms into C code is a challenging task that is not only time-consuming but also error-prone due to the many inherent differences between the two languages. Automatic generation of C/C++ code from algorithms written in MATLAB can accelerate development processes and increase quality, essentially saving time and cost.
In this session, MathWorks experts demonstrate the workflow for generating readable and portable C/C++ code from algorithms written in MATLAB using MATLAB Coder. The session covers various use cases for the generated code, including generating a standalone compiled executable, a library that can be shared and integrated, or MEX-functions for verification or for accelerating simulations. MATLAB Coder supports the use of System objects in DSP System Toolbox, Communications System Toolbox, and Computer Vision System Toolbox for code generation, thus completing the workflow for modeling, simulating, and implementing complex systems from within MATLAB.
The session Design of Signal and Video Processing Systems Using System Objects details the use of System objects.
Knowledge-based modeling is more in demand as the need for more advanced and complex systems arises. Consumers want more value from products. Current available standard tools are not able to handle the design and analysis requirement due to the fixed boundary of their theory and the architecture. Knowledge-based modeling not only can fill this design and analysis gap but also makes the overall innovation and initial tradeoffs much faster.
Using MATLAB, Whirlpool is creating a whole new way for designing a product, from idea to production. The complete process is automated; the requirements are input and the design is optimized for performance of the system, different ideas, cost, and other parameters of interest. The presentation includes an example of refrigeration design to show Whirlpool's philosophy.
Puneet Kumar, MathWorks
Teams developing complex applications for implementation on field-programmable gate arrays (FPGAs) face increasing pressure to verify their designs early and find errors before reaching the final implementation stage. In this session, you'll discover how you can reduce the time it takes to design complex signal processing and communications systems on FPGAs, and streamline the process of verifying and implementing your designs using Simulink HDL Coder™, EDA Simulator Link™, and other MathWorks products. MathWorks experts demonstrate tool capabilities such as the HDL Workflow Advisor and FPGA-in-the-loop (FIL) verification that help you automate and streamline the entire workflow for implementing and verifying designs on FPGA platforms.
Shobhit Shanker, MathWorks
Verification and validation techniques applied throughout the development process enable you to find errors before they can derail your project. This presentation, supported by example models, illustrates the latest techniques and demonstrates the effectiveness of performing activities such as closed-loop simulation. We discuss model-to-software verification techniques such as software-in-the-loop (SIL) and processor-in-the-loop (PIL) testing that can be applied to handwritten or model-generated code to confirm that the behaviour of the software matches the behaviour of the model. We also show solutions you can use to move from desktop environments to real-time simulation of models on target computers for rapid control prototyping, hardware-in-the-loop (HIL) simulation, and other applications. Additionally, the benefits of HIL and real-time simulation, such as minimising costs and risks in embedded system development before deployment in production, are discussed.
Chirag Patel, MathWorks
The tasks of designing, prototyping, and testing electronics systems have become increasingly complex. Flexible analysis tools and modern measurement hardware are needed to meet today's design and testing challenges. You can use MATLAB with Data Acquisition Toolbox™, Image Acquisition Toolbox™, Instrument Control Toolbox™, and MATLAB Report Generator™ and other associated products to acquire and analyse data, build applications, generate reports, and automate measurements. In this session, MathWorks experts demonstrate how to make custom measurements, perform detailed data analysis, configure and control hardware, and develop interactive as well as automated measurement and testing systems. Learn how you can use MATLAB to create efficient test and measurement systems and processes.
Simulink and Stateflow from MathWorks are the dominant choice in the industry for building automotive controller models at the early stages of system development. Effective functional testing at this stage can result in high-quality, error-free reference models resulting in large cost savings. This testing can be done using formal verifications tools such as Simulink Design Verifier™ that exhaustively verify the models against their requirements. In this session, we describe a methodology for applying such tools and share our experience through a number of case studies.
Tom Erkkinen, MathWorks
The increasing prevalence of safety standards, including DO-178, DO-254, ISO 26262, and IEC 61508, is forcing organisations to re-evaluate strategies for system verification and validation. This in turn is leading organisations to adopt Model-Based Design for system design. This presentation showcases a workflow that is used throughout requirements validation, system design, implementation, and testing. It shows how different MathWorks products such as Simulink Design Verifier™ and Polyspace® code verifiers achieve a workflow that caters to different safety-critical standards. Specifically, techniques for establishing traceability, ensuring conformance to design standards, and verifying the output of each design stage are highlighted. The session also details various safety-critical objectives you can achieve by following the demonstrated workflow.
Traditional methods of tuning key powertrain parameters involve experimenting with a couple parameters at a time on prototype vehicles. This method is time-consuming and expensive and may leave room for further optimisation in the vehicle design to meet emission regulations and performance requirements. In this master class, we showcase how simulation-led design optimisation helps you reduce the time and cost involved in optimising system-level performance. Topics include:
- Building fast and accurate multidomain system-level simulations by leveraging measured data
- Speeding up simulations and optimisation runs using parallel computing
- Automatically tuning model parameters using optimisation algorithms
Though the case study pertains to an automotive application, the methods covered in this session can be applied to many system design and optimisation problems.